Review: Strategies for using satellite-based products in modeling PM2.5 and short-term pollution episodes

Evaluating low-cost monitoring designs for PM2.5 exposure assessment with a spatiotemporal modeling approach

Determining the most feasible and cost-effective approaches to improving PM2.5 exposure assessment with low-cost monitors (LCMs) can considerably enhance the quality of its epidemiological inferences. We investigated features of fixed-site LCM designs that most impact PM2.5 exposure estimates to be used in long-term epidemiological inference for the Adult Changes in Thought Air Pollution (ACT-AP) study. We used ACT-AP collected and calibrated LCM PM2.5 measurements at the two-week level from April 2017 to September 2020 (N of monitors [measurements] = 82 [502]). We also acquired reference-grade PM2.5 measurements from January 2010 to September 2020 (N = 78 [6186]). We used a spatiotemporal modeling approach to predict PM2.5 exposures with either all LCM measurements or varying subsets with reduced temporal or spatial coverage. We evaluated the models based on a combination of cross-validation and external validation at locations of LCMs included in the models (N = 82), and also based on an independent external validation with a set of LCMs not used for the modeling (N = 30). We found that the model's performance declined substantially when LCM measurements were entirely excluded (spatiotemporal validation R2 [RMSE] = 0.69 [1.2 μg/m3]) compared to the model with all LCM measurements (0.84 [0.9 μg/m3]). Temporally, using the farthest apart measurements (i.e., the first and last) from each LCM resulted in the closest model's performance (0.79 [1.0 μg/m3]) to the model with all LCM data. The models with only the first or last measurement had decreased performance (0.77 [1.1 μg/m3]). Spatially, the model's performance decreased linearly to 0.74 (1.1 μg/m3) when only 10% of LCMs were included. Our analysis also showed that LCMs located in densely populated, road-proximate areas improved the model more than those placed in moderately populated, road-distant areas.

A comprehensive review of the development of land use regression approaches for modeling spatiotemporal variations of ambient air pollution: A perspective from 2011 to 2023

Land use regression (LUR) models are widely used in epidemiological and environmental studies to estimate humans' exposure to air pollution within urban areas. However, the early models, developed using linear regressions and data from fixed monitoring stations and passive sampling, were primarily designed to model traditional and criteria air pollutants and had limitations in capturing high-resolution spatiotemporal variations of air pollution. Over the past decade, there has been a notable development of multi-source observations from low-cost monitors, mobile monitoring, and satellites, in conjunction with the integration of advanced statistical methods and spatially and temporally dynamic predictors, which have facilitated significant expansion and advancement of LUR approaches. This paper reviews and synthesizes the recent advances in LUR approaches from the perspectives of the changes in air quality data acquisition, novel predictor variables, advances in model-developing approaches, improvements in validation methods, model transferability, and modeling software as reported in 155 LUR studies published between 2011 and 2023. We demonstrate that these developments have enabled LUR models to be developed for larger study areas and encompass a wider range of criteria and unregulated air pollutants. LUR models in the conventional spatial structure have been complemented by more complex spatiotemporal structures. Compared with linear models, advanced statistical methods yield better predictions when handling data with complex relationships and interactions. Finally, this study explores new developments, identifies potential pathways for further breakthroughs in LUR methodologies, and proposes future research directions. In this context, LUR approaches have the potential to make a significant contribution to future efforts to model the patterns of long- and short-term exposure of urban populations to air pollution.

Maternal exposure to particulate matter early in pregnancy and congenital anomalies in offspring: Analysis of concentration-response relationships in a population-based cohort with follow-up throughout childhood

Studies have suggested an association between particulate matter (PM) air pollution and certain congenital anomalies (CAs). However, most studies assumed a linear concentration-response relation and were based on anomalies that were ascertained at birth or up to 1 year of age. We investigated associations between exposures to PM during the first trimester of pregnancy and CAs in 9 organ systems using birth and childhood follow-up data from a leading health care provider in Israel. We conducted a retrospective population-based cohort study among 396,334 births, 2004-2015. Daily PM data at a 1 × 1 km spatial grid were obtained from a satellite-derived prediction models and were linked to the mothers' residential addresses at birth. Adjusted odds ratios (ORs) were estimated with logistic regression models using exposure levels as either continuous or categorical variables. We captured 57,638 isolated CAs with estimated prevalence of 96 and 136 anomalies per 1000 births in the first year of life and by age 6 years, respectively. Analysis of continuous PM with diameter < 2.5 μm (PM_2.5) indicated a supra-linear relation with anomalies in the circulatory, respiratory, digestive, genital and integument systems (79 % of CAs). The slope of the concentration-response function was positive and steepest for PM_2.5 lower than the median concentration (21.5 μg/m³) and had a less steep or negative slope at higher levels. Similar trends were observed for PM_2.5 quartiles. For example, for cardiac anomalies, the ORs were 1.09 (95 % confidence interval: 1.02, 1.15), 1.04 (0.98, 1.10) and 1.00 (0.94, 1.07) for births in the second, third and fourth quartiles, respectively, when compared to the first quartile. In sum, this study adds new evidence for adverse effects of air pollution on neonatal health even with low-level air pollution. Information on late diagnosis of children with anomalies is important in evaluating the burden of disease.

A systematic review and meta-analysis of air pollution and angina pectoris attacks: identification of hazardous pollutant, short-term effect, and vulnerable population

We conducted a systematic review and meta-analysis of global epidemiological studies of air pollution and angina pectoris, aiming to explore the deleterious air pollutant(s) and vulnerable sub-populations. PubMed and Web of Science databases were searched for eligible articles published between database inception and October 2021. Meta-analysis weighted by inverse-variance was utilized to pool effect estimates based on the type of air pollutant, including particulate matters (PM_2.5 and PM₁₀: particulate matter with an aerodynamic diameter ≤ 2.5 µm and ≤ 10 µm), gaseous pollutants (NO₂: nitrogen dioxide; CO: carbon monoxide; SO₂: sulfur dioxide, and O₃: ozone). Study-specific effect estimates were standardized and calculated with percentage change of angina pectoris for each 10 µg/m³ increase in air pollutant concentration. Twelve studies involving 663,276 angina events from Asia, America, Oceania, and Europe were finally included. Meta-analysis showed that each 10 µg/m³ increase in PM_2.5 and PM₁₀ concentration was associated with an increase of 0.66% (95%CI: 0.58%, 0.73%; p < 0.001) and 0.57% (95%CI: 0.20%, 0.94%; p = 0.003) in the risk of angina pectoris on the second day of exposure. Adverse effects were also observed for NO₂ (0.67%, 95%CI: 0.33%, 1.02%; p < v0.001) on the second day, CO (0.010%, 95%CI: 0.006%, 0.014%; p < 0.001). The elderly and patients with coronary artery disease (CAD) appeared to be at higher risk of angina pectoris. Our findings suggest that short-term exposure to PM_2.5, PM₁₀, NO₂, and CO was associated with an increased risk of angina pectoris, which may have implications for cardiologists and patients to prevent negative cardiovascular outcomes.

Does It Measure Up? A Comparison of Pollution Exposure Assessment Techniques Applied across Hospitals in England

Weighted averages of air pollution measurements from monitoring stations are commonly assigned as air pollution exposures to specific locations. However, monitoring networks are spatially sparse and fail to adequately capture the spatial variability. This may introduce bias and exposure misclassification. Advanced methods of exposure assessment are rarely practicable in estimating daily concentrations over large geographical areas. We propose an accessible method using temporally adjusted land use regression models (daily LUR). We applied this to produce daily concentration estimates for nitrogen dioxide, ozone, and particulate matter in a healthcare setting across England and compared them against geographically extrapolated measurements (inverse distance weighting) from air pollution monitors. The daily LUR estimates outperformed IDW. The precision gains varied across air pollutants, suggesting that, for nitrogen dioxide and particulate matter, the health effects may be underestimated. The results emphasised the importance of spatial heterogeneity in investigating the societal impacts of air pollution, illustrating improvements achievable at a lower computational cost.

Exploring the detailed spatiotemporal characteristics of PM2.5: Generating a full-coverage and hourly PM2.5 dataset in the Sichuan Basin, China

Fine particulate matter (PM_2.5) has received worldwide attention due to its threat to public health. In the Sichuan Basin (SCB), PM_2.5 is causing heavy health burdens due to its high concentrations and population density. Compared with other heavily polluted areas, less effort has been made to generate a full-coverage PM_2.5 dataset of the SCB, in which the detailed PM_2.5 spatiotemporal characteristics remain unclear. Considering commonly existing spatiotemporal autocorrelations, the top-of-atmosphere reflectance (TOAR) with a high coverage rate and other auxiliary data were employed to build commonly used random forest (RF) models to generate accurate hourly PM_2.5 concentration predictions with a 0.05° × 0.05° spatial resolution in the SCB in 2016. Specifically, with historical concentrations predicted from a spatial RF (S-RF) and observed at stations, an alternative spatiotemporal RF (AST-RF) and spatiotemporal RF (ST-RF) were built in grids with stations (type 1). The predictions from the AST-RF in grids without stations (type 2) and observations in type 1 formed the PM_2.5 dataset. The LOOCV R², RMSE and MAE were 0.94/0.94, 8.71/8.62 μg∕m³ and 5.58/5.57 μg∕m³ in the AST-RF/ST-RF, respectively. Using the produced dataset, spatiotemporal analysis was conducted for a detailed understanding of the spatiotemporal characteristics of PM_2.5 in the SCB. The PM_2.5 concentrations gradually increased from the edge to the center of the SCB in spatial distribution. Two high-concentration areas centered on Chengdu and Zigong were observed throughout the year, while another high-concentration area centered on Dazhou was only observed in winter. The diurnal variation had double peaks and double valleys in the SCB. The concentrations were high at night and low in daytime, which suggests that characterizing the relationship between PM_2.5 and adverse health outcomes by daily means might be inaccurate with most human activities conducted in daytime.

Model development and validation of personal exposure to PM2.5 among urban elders

Indirect measurements through a combination of microenvironment concentrations and personal activity diaries provide a potentially useful alternative for PM_2.5 exposure estimates. This study was to optimize a personal exposure model based on spatiotemporal model predictions for PM_2.5 exposure in a sub-cohort study. Personal, home indoor, home outdoor, and ambient monitoring data of PM_2.5 were conducted for an elderly population in the Taipei city of Taiwan. The proposed microenvironment exposure (ME) models incorporate PM_2.5 measurements and individual time-activity information with a generalized estimating equation (GEE) analysis. We evaluated model performance with daily personal PM_2.5 exposure based on the coefficient of determination, accuracy, and mean bias error. Ambient and home outdoor measures as exposure surrogates are likely to under- and overestimate personal exposure to PM_2.5 in our study population, respectively. Measured and predicted indoor exposures were highly correlated with personal PM_2.5 exposure. The awareness of peculiar smells is an important factor that significantly increases personal PM_2.5 exposure by 46-70%. The model incorporating home indoor PM_2.5 can achieve the highest agreement (R² = 0.790) with personal exposure and the lowest measurement error. The ME model with the GEE analysis combining home outdoor PM_2.5 determined by LUR model with a machine learning technique can improve the prediction (R² = 0.592) of personal PM_2.5 exposure, compared with the prediction of the traditional LUR model (R² = 0.385).

Avoidable mortality due to long-term exposure to PM2.5 in Colombia 2014-2019

OBJECTIVE

To compare estimates of spatiotemporal variations of surface PM_2.5 concentrations in Colombia from 2014 to 2019 derived from two global air quality models, as well as to quantify the avoidable deaths attributable to the long-term exposure to concentrations above the current and projected Colombian standard for PM_2.5 annual mean at municipality level.

METHODS

We retrieved PM_2.5 concentrations at the surface level from the ACAG and CAMSRA global air quality models for all 1,122 municipalities, and compare 28 of them with available concentrations from monitor stations. Annual mortality data 2014-2019 by municipality of residence and pooled effect measures for total, natural and specific causes of mortality were used to calculate the number of annual avoidable deaths and years of potential life lost (YPLL) related to the excess of PM_2.5 concentration over the current mean annual national standard of 25 µg/m³ and projected standard of 15 µg/m³.

RESULTS

Compared to surface data from 28 municipalities with monitoring stations in 2019, ACAG and CAMSRA models under or overestimated annual mean PM_2.5 concentrations. Estimations from ACAG model had a mean bias 1,7 µg/m³ compared to a mean bias of 4,7 µg/m³ from CAMSRA model. Using ACAG model, estimations of total nationally attributable deaths to PM_2.5 exposure over 25 and 15 µg/m³ were 142 and 34,341, respectively. Cardiopulmonary diseases accounted for most of the attributable deaths due to PM_2.5 excess of exposure (38%). Estimates of YPLL due to all-cause mortality for exceeding the national standard of 25 µg/m³ were 2,381 years.

CONCLUSION

Comparison of two global air quality models for estimating surface PM_2.5 concentrations during 2014-2019 at municipality scale in Colombia showed important differences. Avoidable deaths estimations represent the total number of deaths that could be avoided if the current and projected national standard for PM_2.5 annual mean have been met, and show the health-benefit of the implementation of more restrictive air quality standards.

Tracking short-term health impacts attributed to ambient PM2.5 and ozone pollution in Chinese cities: an assessment integrates daily population

Joint and synergistic control of PM_2.5 and ozone pollution is an urgent need in China and a global-widely concerned issue. Health impact assessment could provide a comprehensive perspective for PM_2.5-ozone coordinated control strategies. For a detailed understanding of the seasonality and regionality of the health impacts attributed to PM_2.5 and ozone in China, this study extended the classic health impact function by daily population and assessed the short-term (daily) health impacts in 335 Chinese cities in 2021. Population migration indexes from Baidu were introduced to estimate the cities' daily population. Using this method, we quantitatively investigated the influence of population on short-term health impact assessment and identified which was significant in the Pearl River Delta (PRD) region and other populous cities. Although the annual sums of PM_2.5- and ozone-related daily health impacts were close for all Chinese cities, the PM_2.5-related health impact was equivalent to 333.96% and 32.07% of that ozone-related, during the cold and warm periods. The correlation and local spatial association analysis found significant city-specific and city-cluster associations of daily health impacts during the warm period and in Beijing-Tianjin-Hebei and surrounding regions (BTHS) and the Yangtze River Delta (YRD). Policymakers could promote period- and pollutant-targeted control actions for the major city groups, especially the BTHS, YRD, and PRD. Our methods and findings investigated the various influences of the population on short-term health impact assessment and proposed the PM_2.5-ozone collaborative control idea for key regions and city groups.

Prediction of daily mean and one-hour maximum PM2.5 concentrations and applications in Central Mexico using satellite-based machine-learning models

BACKGROUND

Machine-learning algorithms are becoming popular techniques to predict ambient air PM_2.5 concentrations at high spatial resolutions (1 × 1 km) using satellite-based aerosol optical depth (AOD). Most machine-learning models have aimed to predict 24 h-averaged PM_2.5 concentrations (mean PM_2.5) in high-income regions. Over Mexico, none have been developed to predict subdaily peak levels, such as the maximum daily 1-h concentration (max PM_2.5).

OBJECTIVE

Our goal was to develop a machine-learning model to predict mean PM_2.5 and max PM_2.5 concentrations in the Mexico City Metropolitan Area from 2004 through 2019.

METHODS

We present a new modeling approach based on extreme gradient boosting (XGBoost) and inverse-distance weighting that uses AOD, meteorology, and land-use variables. We also investigated applications of our mean PM_2.5 predictions that can aid local authorities in air-quality management and public-health surveillance, such as the co-occurrence of high PM_2.5 and heat, compliance with local air-quality standards, and the relationship of PM_2.5 exposure with social marginalization.

RESULTS

Our models for mean and max PM_2.5 exhibited good performance, with overall cross-validated mean absolute errors (MAE) of 3.68 and 9.20 μg/m³, respectively, compared to mean absolute deviations from the median (MAD) of 8.55 and 15.64 μg/m³. In 2010, everybody in the study region was exposed to unhealthy levels of PM_2.5. Hotter days had greater PM_2.5 concentrations. Finally, we found similar exposure to PM_2.5 across levels of social marginalization.

SIGNIFICANCE

Machine learning algorithms can be used to predict highly spatiotemporally resolved PM_2.5 concentrations even in regions with sparse monitoring.

IMPACT

Our PM_2.5 predictions can aid local authorities in air-quality management and public-health surveillance, and they can advance epidemiological research in Central Mexico with state-of-the-art exposure assessment methods.

Revisiting the atmospheric particles: Connecting lines and changing paradigms

Historically, the atmospheric particles constitute the most primitive and recent class of air pollutants. The science of atmospheric particles erupted more than a century ago covering more than four decades of size, with past few years experiencing major advancements on both theoretic and data-based observational grounds. More recently, the plausible recognition between particulate matter (PM) and the diffusion of the COVID-19 pandemic has led to the accretion of interest in particle science. With motivation from diverse particle research interests, this paper is an 'old engineer's survey' beginning with the evolution of atmospheric particles and identifies along the way many of the global instances signaling the 'size concept' of PM. A theme that runs through the narrative is a 'previously known' generational evolution of particle science to the 'newly procured' portfolio of knowledge, with important gains on the application of unmet concepts and future approaches to PM exposure and epidemiological research.

Ensemble averaging using remote sensing data to model spatiotemporal PM10 concentrations in sparsely monitored South Africa

There is a paucity of air quality data in sub-Saharan African countries to inform science driven air quality management and epidemiological studies. We investigated the use of available remote-sensing aerosol optical depth (AOD) data to develop spatially and temporally resolved models to predict daily particulate matter (PM₁₀) concentrations across four provinces of South Africa (Gauteng, Mpumalanga, KwaZulu-Natal and Western Cape) for the year 2016 in a two-staged approach. In stage 1, a Random Forest (RF) model was used to impute Multiangle Implementation of Atmospheric Correction AOD data for days where it was missing. In stage 2, the machine learner algorithms RF, Gradient Boosting and Support Vector Regression were used to model the relationship between ground-monitored PM₁₀ data, AOD and other spatial and temporal predictors. These were subsequently combined in an ensemble model to predict daily PM₁₀ concentrations at 1 km × 1 km spatial resolution across the four provinces. An out-of-bag R² of 0.96 was achieved for the first stage model. The stage 2 cross-validated (CV) ensemble model captured 0.84 variability in ground-monitored PM₁₀ with a spatial CV R² of 0.48 and temporal CV R² of 0.80. The stage 2 model indicated an optimal performance of the daily predictions when aggregated to monthly and annual means. Our results suggest that a combination of remote sensing data, chemical transport model estimates and other spatiotemporal predictors has the potential to improve air quality exposure data in South Africa's major industrial provinces. In particular, the use of a combined ensemble approach was found to be useful for this area with limited availability of air pollution ground monitoring data.

Air Pollution, Oxidative Stress, and the Risk of Development of Type 1 Diabetes

Despite multiple studies focusing on environmental factors conducive to the development of type 1 diabetes mellitus (T1DM), knowledge about the involvement of long-term exposure to air pollution seems insufficient. The main focus of epidemiological studies is placed on the relationship between exposure to various concentrations of particulate matter (PM): PM1, PM2.5, PM10, and sulfur dioxide (SO2), nitrogen dioxide (NO2), carbon monoxide (O3), versus the risk of T1DM development. Although the specific molecular mechanism(s) behind the link between increased air pollution exposure and a higher risk of diabetes and metabolic dysfunction is yet unknown, available data indicate air pollution-induced inflammation and oxidative stress as a significant pathway. The purpose of this paper is to assess recent research examining the association between inhalation exposure to PM and associated metals and the increasing rates of T1DM worldwide. The development of modern and more adequate methods for air quality monitoring is also introduced. A particular emphasis on microsensors, mobile and autonomous measuring platforms, satellites, and innovative approaches of IoT, 5G connections, and Block chain technologies are also presented. Reputable databases, including PubMed, Scopus, and Web of Science, were used to search for relevant literature. Eligibility criteria involved recent publication years, particularly publications within the last five years (except for papers presenting a certain novelty or mechanism for the first time). Population, toxicological and epidemiological studies that focused particularly on fine and ultra-fine PM and associated ambient metals, were preferred, as well as full-text publications.

Spatiotemporal variations in aerosol optical depth and associated risks for populations in the arid region of Central Asia

With the progress of urbanization, atmospheric pollution and physical health issues caused by the increase of aerosol optical depth (AOD) become more and more prominent. Hence, population exposure risk to AOD becomes a research hotspot. The arid Central Asia (ACA) has a generally high AOD and is a major source area for dust aerosols in the world. Only few studies have discussed population exposure risk to AOD in ACA. Based on multisource remote sensing data, and used population exposure risk model, this study evaluated population exposure risk to AOD in six ecological zones (Northern steppe region of ACA (NSCA), Aral Sea desert area (ASDA), Tianshan Mountains (TSMT), Junggar Basin desert area (JBDA), Tarim Basin desert area (TBDA) and Hexi corridor desert area (HCDA)). Generally, AOD in ACA was kept increasing from 2000 to 2015, and it increased mostly in HCDA and areas near the Aral Sea (p < 0.001). With respect to seasonal variations, the maximum AOD was observed in spring and autumn, and the minimum was in winter. Considering land use changes, AOD was mainly manifested by the reduction of water bodies and expansion of construction lands. This was the mostly significant in NSCA and ASDA (p < 0.01). The population exposure risk to AOD in ACA was increasing continuously from 2000 to 2015, and high-value regions (>9) concentrated in oases, specifically, in the Aral Sea basin and Tarim River basin.The Aral Sea basin became the major AOD source region in ACA due to the shrinking water area after unreasonable development and utilization of water resources. These further increase population exposure risk to AOD in the Aral Sea area. Hence, ecological restoration in terminal lakes of ACA will become the key to lower population exposure risk to AOD practically.

Spatial and Temporal Analysis of Impacts of Hurricane Florence on Criteria Air Pollutants and Air Toxics in Eastern North Carolina

Natural and anthropogenic disasters are associated with air quality concerns due to the potential redistribution of pollutants in the environment. Our objective was to conduct a spatiotemporal analysis of air concentrations of benzene, toluene, ethylbenzne, and xylene (BTEX) and criteria air pollutants in North Carolina during and after Hurricane Florence. Three sampling campaigns were carried out immediately after the storm (September 2018) and at four-month intervals. BTEX were measured along major roads. Concurrent criteria air pollutant concentrations were predicted from modeling. Correlation between air pollutants and possible point sources was conducted using spatial regression. Exceedances of ambient air criteria were observed for benzene (in all sampling periods) and PM2.5 (mostly immediately after Florence). For both, there was an association between higher concentrations and fueling stations, particularly immediately after Florence. For other pollutants, concentrations were generally below levels of regulatory concern. Through characterization of air quality under both disaster and "normal" conditions, this study demonstrates spatial and temporal variation in air pollutants. We found that only benzene and PM2.5 were present at levels of potential concern, and there were localized increases immediately after the hurricane. These substances warrant particular attention in future disaster response research (DR2) investigations.

Spatiotemporal Dynamics of Ecological Security Pattern of Urban Agglomerations in Yangtze River Delta Based on LUCC Simulation

Urbanization has not only promoted economic development, but also significantly changed land use and development strategy. The environmental problems brought by urbanization threaten ecological security directly. Therefore, it is necessary to introduce changes in land use when constructing an ecological security pattern. This study takes the Yangtze River Delta urban agglomeration, one of the most economically developed regions in China, as the research area. Based on its land use status, the Cellular Automata–Markov model was used to predict the quantitative change and transfer of land-use types in 2025, and three types of land-use patterns were simulated under different scenarios. Combined with the pressure–state–response model, the Entropy TOPSIS comprehensive evaluation model is used to evaluate the three phases in the years of 2005, 2010, and 2015, and the results indicated that the safety level dropped from 85.45% to 82.94%. Five spatial associations were obtained from the spatial autocorrelation analysis using GeoDA, and the clustering distribution of the three phases was roughly the same. Based on the requirements of “Natural Growth” scenario, “Urban Sprawl” scenario, and “Ecological Protection” scenario, the transfer matrix of the various land-use types were modified rationally. The results of scenario simulations illustrated that the level of urbanization was inversely proportional to the level of ecological security. The surrounding cities in the northern part of Taihu Lake were developing rapidly, with low levels of ecological security. The hilly cities in the southern part, in contrast, developed slowly and had a high level of ecological security. Based on the temporal and spatial changes in the ecosystem, an ecosystem optimization model was proposed to determine the ecological functional areas. The nature of each functional area provided the basis to formulate urban construction and management plans and achieve sustainable urban development.

A machine learning model to estimate ambient PM2.5 concentrations in industrialized highveld region of South Africa

Exposure to fine particulate matter (PM_2.5) has been linked to a substantial disease burden globally, yet little has been done to estimate the population health risks of PM_2.5 in South Africa due to the lack of high-resolution PM_2.5 exposure estimates. We developed a random forest model to estimate daily PM_2.5 concentrations at 1 km² resolution in and around industrialized Gauteng Province, South Africa, by combining satellite aerosol optical depth (AOD), meteorology, land use, and socioeconomic data. We then compared PM_2.5 concentrations in the study domain before and after the implementation of the new national air quality standards. We aimed to test whether machine learning models are suitable for regions with sparse ground observations such as South Africa and which predictors played important roles in PM_2.5 modeling. The cross-validation R² and Root Mean Square Error of our model was 0.80 and 9.40 μg/m³, respectively. Satellite AOD, seasonal indicator, total precipitation, and population were among the most important predictors. Model-estimated PM_2.5 levels successfully captured the temporal pattern recorded by ground observations. Spatially, the highest annual PM_2.5 concentration appeared in central and northern Gauteng, including northern Johannesburg and the city of Tshwane. Since the 2016 changes in national PM_2.5 standards, PM_2.5 concentrations have decreased in most of our study region, although levels in Johannesburg and its surrounding areas have remained relatively constant. This is anadvanced PM_2.5 model for South Africa with high prediction accuracy at the daily level and at a relatively high spatial resolution. Our study provided a reference for predictor selection, and our results can be used for a variety of purposes, including epidemiological research, burden of disease assessments, and policy evaluation.

Grey Correlation Analysis of Haze Impact Factor PM2.5

In recent years, frequent severe haze weather has formed in China, including some of the most populated areas. We found that these smog-prone areas are often relatively a “local climate” and aim to explore this series of scientific problems. This paper uses remote sensing and data mining methods to study the correlation between haze weather and local climate. First, we select Beijing, China and its surrounding areas (East longitude 115°20′11″–117°40′35″, North latitude 39°21′11″–41°7′51″) as the study area. We collected data from meteorological stations in Beijing and Xianghe from March 2014 to February 2015, and analyzed the meteorological parameters through correlation analysis and a grey correlation model. We study the correlation between the six influencing factors of temperature, dew point, humidity, wind speed, air pressure and visibility and PM2.5, so as to analyze the correlation between haze weather and local climate more comprehensively. The results show that the influence of each index on PM2.5 in descending order is air pressure, wind speed, humidity, dew point, temperature and visibility. The qualitative analysis results confirm each other. Among them, air pressure (correlation 0.771) has the greatest impact on haze weather, and visibility (correlation 0.511) is the weakest.